Create a RasterLayer Object from Mask or Dsurface
Methods to convert secr object to a RasterLayer object.
## S4 method for signature 'mask' raster(x, covariate, values = 1, crs = NA) ## S4 method for signature 'Dsurface' raster(x, covariate, values = 1, crs = NA)
x |
mask or Dsurface object |
covariate |
character name of covariate to provide values for RasterLayer |
values |
numeric values for RasterLayer |
crs |
character or object of class CRS. Optional PROJ.4 type description of a Coordinate Reference System (map projection). |
There are two ways to specify the values to be used. If covariate
is provided then
the values of the corresponding covariate of the mask or Dsurface are used. Otherwise,
values
is duplicated to the required number of rows.
The resulting RasterLayer may optionally include a PROJ.4 map projection
defined via crs
. The specification may be very simple (as in the
example below) or complex, including an explicit datum and other
arguments. Projections are used by raster, sp, rgdal
and other packages. See raster
for further
explanation and links.
The S3 classes ‘mask’ and ‘Dsurface’ are defined in secr as virtual S4 classes. This enables these extensions to the list of S4 methods defined in raster.
Although these methods work ‘standalone’, it is currently necessary to load the raster package to do much with the result (e.g., plot it).
RasterLayer
Prior to secr 2.9.5 these methods could fail unpredictably because an intermediate array was badly dimensioned due to truncation of a floating point value.
## Not run: shorePossums <- predictDsurface(possum.model.Ds) tmp <- raster(shorePossums, covariate = "D.0") library(raster) plot(tmp, useRaster = FALSE) ## alternative with same result tmp <- raster(shorePossums, values = covariates(shorePossums)$D.0) ## set the projection ## here the crs PROJ.4 spec refers simply to the old NZ metric grid tmp <- raster(shorePossums, "D.0", crs = "+proj=nzmg") ## check the projection proj4string(tmp) ## End(Not run)
Please choose more modern alternatives, such as Google Chrome or Mozilla Firefox.